بازشناسی عادی شباهت عاطفی بین حالات چهره پس از آسیب آمیگدال دوجانبه

Abstract Bilateral damage to the amygdala in humans has been previously linked to two deficits in recognizing emotion in facial expressions: recognition of individual basic emotions, especially fear, and recognition of similarity among emotional expressions. Although several studies have examined recognition of individual emotions following amygdala damage, only one subject has been examined on recognition of similarity. To assess the extent to which deficits in recognizing similarity among facial expressions might be a general consequence of amygdala damage, we examined this ability in two subjects with complete bilateral amygdala damage. Both subjects had previously demonstrated entirely normal recognition of individual facial emotions. Here we report that these two patients also are intact in their ability to recognize similarity between emotional expressions. These results indicate that, like the recognition of individual basic emotions in facial expressions, the recognition of similarity among emotional expressions does not have an absolute dependence on the amygdala.

Introduction Several recent neuropsychological studies have investigated the role of the amygdala in the recognition of emotion in human facial expressions. The first of these studies described patient S.M., a 30-year old female with nearly complete bilateral destruction of the amygdala through calcification resulting from congenital Urbach-Wiethe disease [2]. In that initial report, two facial emotion processing deficits were identified in S.M. The first was an impairment in recognizing specific emotions in facial expressions, particularly fear. This deficit was revealed by her inability to recognize the emotion fear in a task that involved judging the intensity of individual basic emotions in facial expressions. The second deficit involved the ability to recognize similarity between different facial expressions of emotion that normal subjects rate as similar. When shown two facial expressions and asked to rate their emotional similarity, S.M. gave abnormally low similarity ratings to pairs of expressions that normal subjects judge as similar, such as anger and disgust or fear and surprise. Further analysis indicated that S.M. gave abnormal similarity ratings because she could only recognize the most dominant or prototypical emotion in each facial expression, in contrast to normal subjects who can recognize that a given facial expression can contain a blend of different intensities of different emotions [17]. From these data, Adolphs et al. [2] concluded that the amygdala may be indispensable to recognizing individual basic emotions in facial expressions (especially fear) and in recognizing similarity between different facial expressions. Following upon this initial report, several neuropsychological studies have investigated deficits in recognizing individual basic emotions in facial expressions in patients with bilateral amygdala damage [4], [6], [22] and [23]. These studies have generally supported the link between bilateral amygdala damage and facial emotion deficits involving fear and other negative emotions, although such deficits are not invariably present [3] and [9]. Neuroimaging studies have also provided evidence that the amygdala has an important role in the perception of fearful, angry, and happy facial expressions [4], [12], [13] and [21]. In contrast to the extensive scrutiny that the relation between amygdala damage and deficits in recognizing individual basic emotions in facial expressions has received, however, no studies have investigated S.M.’s other deficit, the inability to recognize similarity between different facial expressions of emotion. It is important to examine the nature of this deficit further in other patients with bilateral amygdala damage for two reasons. First, it is not known to what extent this facial emotion deficit is general to other cases of bilateral amygdala damage besides S.M. Secondly, it is unclear whether the two facial emotion recognition deficits that S.M. exhibited are separate or whether they simply reflect different aspects of the same underlying deficit. The current study investigated these issues in two postencephalitic patients, E.P. and G.T. [9], each of whom have complete bilateral amygdala lesions. We assessed the ability of E.P. and G.T. to recognize similarity between emotional expressions using the same task that Adolphs et al. [2] used previously with patient S.M. In addition to collecting these new data, we conducted a reanalysis of the raw facial emotion recognition data from a previous study with these patients [9], using multidimensional scaling (MDS) techniques to address the same issue in a different way. In an MDS plot, facial expressions of emotion that a subject perceives as similar are represented as close together, whereas expressions that are perceived as dissimilar are represented as far apart. Patient S.M. exhibited an abnormal MDS plot that indicated that she failed to perceive normally the similarity between different facial expressions of emotion [2]. We compared the MDS plots generated for E.P. and G.T. to those obtained previously for S.M. and normal control subjects to examine whether E.P. and G.T. would be normal or abnormal on this additional measure. E.P. and G.T. are notable in that they have been shown previously to exhibit normal recognition of specific facial emotions, including fear, despite complete bilateral lesions to the amygdala [9]. Thus, if E.P. and G.T. are impaired on recognizing similarity between different facial expressions, this would imply both that the similarity-recognition deficit is not unique to S.M. and also that this ability is dissociable from the ability to recognize specific emotions in facial expressions, suggesting that the two facial emotion deficits that S.M. previously exhibited are in fact separate. Conversely, if both E.P. and G.T. can recognize similarity between emotional expressions normally, this would indicate that the deficit that S.M. exhibited in this ability does not generalize to all cases of bilateral amygdala damage and would demonstrate that this ability does not have an absolute dependence on the amygdala, paralleling the findings of the previous report that examined the ability of these two patients to recognize specific emotions in facial expressions [9

Results 3.1. Facial expression similarity task Fig. 1 shows the results for the facial expression similarity task for patients E.P. and G.T. for facial expression pairs in which the two faces depicted different emotions. Each category shows average similarity ratings of all 54 possible pairings of expressions (three expressions of the given emotion ×six other emotions ×three expressions of each of these six emotions). The data for E.P. and G.T. are plotted against the data from Adolphs et al. [2] obtained using the same task (data are from two experiments with S.M. and one experiment with eight brain-damaged control subjects). Unlike control subjects, who perceived a moderate degree of similarity between faces expressing different emotions, S.M.’s ratings reflected a tendency not to endorse any emotions other than the protypical or dominant ones depicted in a facial expression. In contrast, patient E.P. clearly exhibited normal performance on this task, giving similarity ratings in the upper range of control performance for every emotion category. G.T.’s performance was also within the performance range of the control subjects for all facial expression categories except two, happiness and fear. G.T.’s similarity ratings for these fell just outside the range of control performance, and G.T.’s scores were also well above both of S.M.’s scores in the category of fear. The one category in which G.T.’s performance matched S.M.’s (happiness) was the one in which control subjects tended to give very low similarity ratings. In summary, G.T.’s overall performance was lower than that of E.P. and lower than any individual control subject, but at the same time his performance was much closer to being completely normal than was S.M.’s. Similarity ratings in the facial expression similarity task for patients E.P. ... Fig. 1. Similarity ratings in the facial expression similarity task for patients E.P. and G.T. (indicated by their initials) for facial expression pairs in which the expressions depicted different emotions. Each category shows average similarity ratings of all 54 possible pairings of expressions (three expressions of the given emotion ×six other emotions ×three expressions of each of these six emotions). All other data points are from Adolphs et al. [2]. Circles indicate the performance of each of eight brain-damaged control subjects whose lesions spared the amygdala. The filled triangles indicate the performance of patient S.M., who was tested twice. Figure options Fig. 2 shows the results for the facial expression similarity task for patients E.P. and G.T. for facial expression pairs in which the two faces depicted the same emotion. Each category shows average similarity ratings of all three possible pairings of the three facial expressions depicting the same emotion. The data for E.P. and G.T. are plotted against the data from Adolphs et al. [2] in the same manner as in Fig. 1. S.M. was not impaired in the ability to recognize similarity between two faces depicting the same emotion, a task which requires only the ability to detect the most prototypical emotion in a facial expression. E.P. and G.T. showed normal performance on this task. Similarity ratings in the facial expression similarity task for patients E.P. ... Fig. 2. Similarity ratings in the facial expression similarity task for patients E.P. and G.T. (indicated by their initials) for facial expression pairs in which the expressions depicted the same emotion. Each category shows average similarity ratings of all three possible pairings of the three facial expressions depicting the same emotion. All other data points are from Adolphs et al. [2]. Circles indicate the performance of each of eight brain-damaged control subjects whose lesions spared the amygdala. The filled triangles indicate the performance of patient S.M., who was tested twice. Figure options 3.2. Multidimensional scaling analysis Fig. 3 shows the MDS plots for the average of seven normal controls (CON), G.T. (GT) and S.M. (SM). For technical reasons related to the MDS procedure, a valid MDS plot could not be computed for the data from E.P. The nearly continuous circular ordering of emotional faces in the MDS plot for G.T. shows that he perceived similarity between the emotional faces normally. For example, in both the CON and GT plots, surprised and afraid faces were adjacent, indicating that they were perceived as being relatively similar. The overall arrangement of emotions in the MDS plots for CON and GT was also similar to what has been found in previous studies of judgments of emotional facial expressions [16] and emotional words [15]. In contrast, S.M.’s abnormal MDS plot showed clustering of the emotional faces into three discrete groups. The gaps in S.M.’s MDS plots corresponded to the endorsement of very low ratings when judging expressions of different emotions on each other’s adjectives (e.g., how much fear is there in surprised expressions, and vice versa?), as illustrated quantitatively in Fig. 1. Multidimensional scaling (MDS) of perceived similarity judgments of emotions ... Fig. 3. Multidimensional scaling (MDS) of perceived similarity judgments of emotions expressed by faces. Each colored point represents one of the 39 facial expressions which subjects rated. Euclidean distance between points corresponds to perceived dissimilarity between simuli. CON=normal control subjects (N=7); GT=patient G.T.; SM=patient S.M.. Data for CON and SM are from Adolphs et al. [2].